Construction / Fintech · Deep dive

Adaptive

AI-native accounting and AP automation for construction — coding every subcontractor invoice to a job and cost code, then trying to run the whole back office with agents.

emerging

The question that decides it: Adaptive's wedge is AP automation that auto-codes every invoice to a job, phase, and cost code; its bet is that AI agents can then run the full month-end close. Does agent-produced job costing get accurate enough that contractors retire Sage 300 CRE or Foundation as their system of record — or does Adaptive stay a capture-and-approval layer bolted onto an ERP it never replaces?

My take ★★ Very interesting ★ Interesting − Maybe ✗ Not interesting

Snapshot

Adaptive is an AI-native accounting and accounts-payable platform built specifically for construction — the contractors, remodelers and homebuilders who run dozens of jobs at once and keep the books in QuickBooks and spreadsheets. Its core act is unglamorous and valuable: it ingests every vendor and subcontractor invoice, codes it to the right job, phase and cost code, routes it for approval, pays it, and syncs the result to the GC’s ERP with little manual entry. Founded in 2021 in New York and launched to market in February 2023, it has raised $26.4 million from Andreessen Horowitz (seed lead, 2022) and Emergence Capital (Series A lead, July 2024), and reports crossing 700 contractor customers by 2026. It is genuinely early — roughly 50 people, no disclosed revenue — and its 2026 pivot to “Project Accounting Agents” is a bet that the same infrastructure can run the entire back office, not just the invoice inbox.

Founding story

The company did not start as construction accounting software. Matt Calvano, Henry Bradlow and Francisco Enriquez were working with a group of homebuilders in Austin on a product for acquiring land and lots. The builders kept redirecting them: the sharper pain was not sourcing dirt, it was the back office — invoices, draws, job costing, and the two-to-three months it routinely takes a construction company to get paid. The founders ran what amounted to a hands-on, white-glove bookkeeping service for builders, saw the fragmentation up close, and concluded the real business was software that could do that work.

The founder mix is the pitch. Calvano came from real-estate finance — Morgan Stanley and PE firm KSL Capital Partners — and understands draws, retainage and the capital stack from the money side. Bradlow wrote guidance algorithms at SpaceX, the credential behind the “AI-native” framing. Enriquez runs finance and payments. All three are Stanford-connected (Calvano finished a Stanford GSB MBA in 2021). They landed a16z as seed lead in July 2022 before writing much code — a bet on the team and the wedge more than the traction.

How it works

Follow one invoice. A subcontractor emails a bill or a PDF lands in the GC’s inbox. Adaptive’s system reads it — OCR plus models — and extracts the vendor, amounts, and line items. The step that matters, and the one generic AP tools skip, is job costing: Adaptive codes the bill not just to an expense account but to a specific job, phase and cost code, the way construction accounting actually has to work. It flags duplicates and likely fraud before anything moves, then routes the bill through an approval workflow with thresholds — small bills auto-approve, larger ones escalate to a project manager or owner.

Once approved, Adaptive pays the vendor (ACH) and can generate the other side of the ledger: client billing in whatever format the contract demands — cost-plus, fixed-price, or AIA-style progress draws — pulling approved job costs with backup attached so a draw package that used to take hours of stapling PDFs comes together in a few clicks. It reconciles payments against bank statements, feeds WIP (work-in-progress) reporting, and syncs the coded transactions back to the customer’s ERP or QuickBooks. In June 2026 Adaptive productized this as “Project Accounting Agents” — software agents that handle AP coding, bill matching, WIP, billings and bank reconciliation, with accountants reviewing exceptions rather than building reports from scratch — reframing the product from tool to labor.

Product and business overview

Adaptive sells modules around the AP core: invoice capture and coding; approvals; vendor payments; client/owner billing and draw management; WIP and job-cost reporting; and bank reconciliation, all sitting on top of (or increasingly in place of) the existing accounting system. The 2026 agent layer is the strategic wrapper — the claim these modules can be run by AI with a controller supervising, aimed as much at outsourced accounting firms as at contractors. A June 2026 partnership with LLUM, an outsourced accounting provider dedicated to construction, embeds Adaptive inside a fractional-CFO service — a channel play that standardizes many GCs’ accountants on Adaptive.

The customer is a general contractor, remodeler or homebuilder running multiple concurrent projects — per the company, businesses with roughly $5M to $1B in revenue — that cannot afford a large accounting staff. Adaptive was named to CB Insights’ 2024 list of the 100 most innovative fintech startups.

Business model and pricing

Adaptive is SaaS, sold on modules rather than pure per-seat licensing, and it makes a point of not charging per user — a jab at incumbents where adding a field approver costs money. Published price points vary by source and have shifted: the pricing page describes a flat Basic plan around $1,000/month with unlimited payments, projects, users and support; other listings (Capterra/Software Finder, accessed 2026) show a revenue-based entry point near $575/month for contractors up to $5M in annual revenue, with quotes scaling from there. The model is explicitly modular — activate AP automation first, add WIP, billing and reconciliation as you grow — with no implementation fee and month-to-month terms.

The honest read: pricing is still being figured out, the two published anchors ($575 and $1,000/month) don’t fully reconcile, and revenue is undisclosed. At roughly 700 customers paying four-figure monthly fees, back-of-envelope ARR is plausibly in the mid-single-digit millions: proof customers want the product, not yet that it can charge enough per customer to justify a venture outcome.

Traction over time

The shape is a fast-growing early-stage company, not a scaled one. From launch in February 2023 to 280+ customers and $1.4B of project volume by July 2024, to 500 and then 700+ through 2025-2026, the customer count is compounding quickly off a small base. The reported NPS of 80 (July 2024) is high and consistent with the review themes below. What is missing is the number that matters for the venture case: net revenue retention and how much a customer expands as it adds modules. Note a common data hazard — several third-party summaries conflate this Adaptive with the identically named deepfake-defense startup Adaptive Security (which raised an $81M Series B in late 2025 to a $146.5M total). They are unrelated.

Market analysis

The construction accounting software market is estimated around $2.5B in 2024, growing at roughly a 6.6% CAGR toward ~$5B by the mid-2030s (Precedence Research, 2025) — a real but not explosive software TAM. The larger framing Adaptive uses is the cost of the problem: the industry loses an estimated $273 billion a year to payment delays, cash-flow crunches and administrative inefficiency (cited in Adaptive’s Series A materials, 2024), and construction firms wait an average of two to three months to get paid. Construction is famously under-digitized, still on paper, PDFs and QuickBooks — the structural tailwind: vertical AI hitting an industry with high labor cost, chronic accountant shortages, and unautomated workflows.

Competitive intel

The competitive set is crowded and better-capitalized on every flank. Procore is the public incumbent that owns the GC’s software relationship and is extending into invoices and pay apps; its threat is bundling, not feature parity. Sage 300 CRE and Sage Intacct Construction are the entrenched systems of record Adaptive integrates with today and hopes to replace tomorrow — decades of job-cost data and CPA habit make them sticky. Foundation Software is the direct construction-accounting incumbent, deep on payroll and union complexity Adaptive does not touch. Knowify fights at the SMB bottom with a cheaper, simpler QuickBooks-adjacent product. The construction-fintech cohort — Siteline ($18.4M, subcontractor billing), Beam ($15M, GC-to-sub payments), and Trimble Pay (the acquired Flashtract) — each attacks a narrower slice of the same money flow, often better-funded on it. And the true default is QuickBooks, which Adaptive replaces customer by customer but which Intuit’s own AI roadmap could defend cheaply.

Where Adaptive wins: it is genuinely construction-native (cost codes, WIP, AIA draws) and modern, against incumbents that are one or the other but rarely both. Where it is exposed: it is the smallest and least-funded serious player in the room, and its edge is a product advantage, not yet a moat.

History and evolution

• 2021 — Founded in New York after pivoting off a land-acquisition idea for Austin homebuilders into their back-office pain; small pre-seed (~$750K) in August.
• Jul 2022 — a16z leads a $6.5M seed; operators from Airbase, Brex and Ramp participate. Product not yet launched.
• Feb 2023 — Platform launches to market.
• 2024 — Named to CB Insights’ 100 Most Innovative Fintech Startups; showcases at Chicago Build Expo.
• Jul 2024 — $19M Series A led by Emergence Capital, a16z following; 280+ customers, $1.4B project volume, NPS 80; total raised $26.4M.
• 2025 — Crosses 500 customers; headcount around 50; positions as “AI-native project accounting.”
• Jun 2026 — Launches Project Accounting Agents (agentic AP coding, WIP, billing, bank rec); partners with outsourced-accounting firm LLUM; reports 700+ contractor customers.

No public crises or layoffs — the company is too young for a real history of stumbles, itself a caveat: the hard parts (scaling payments, support, multi-entity accounting) are still ahead.

What people say

The case for. On Capterra and G2 (2026), construction is ~84% of reviewers, and the recurring praise is concrete: automation kills data entry and frees staff for higher-value work; draw-package generation that used to take hours across 100+ bills now takes a few clicks; support is fast and reachable via chat, often replying within minutes; and the AI that reads emails and turns them into bills and receipts genuinely lands. The July 2024 NPS of 80 is unusually high for accounting software, and construction-dedicated outsourced accountants (LLUM) chose to build a service on top of it.

The complaints. The same reviews name real friction. Users describe the finance-and-accounting depth as not very user-friendly; receipt line-item breakdowns condense data so totals don’t fit on one screen and require scrolling; the reporting function feels clunky; and there are specific gripes about handling vendor credits. Beyond reviews, the bigger honest negatives are structural: this is a ~50-person, $26M-raised company selling into an industry that trusts incumbents and switches slowly; it has no disclosed revenue or retention data; its pricing is still unsettled; and the agent pitch — let AI run the close — collides directly with the fact that accounting errors in job costing are expensive and contractors are conservative. The unrelated-namesake problem (Adaptive Security) also muddies its search footprint and makes diligence harder.

Outlook: the open question

Adaptive works if agent-coded job costing becomes accurate and trusted enough that contractors let it run the close and retire their legacy ERP; it stalls if it stays a capture-and-approval layer that must sync into Sage or Foundation forever. That is the whole thing. The wedge — AP automation that codes invoices to jobs and cost codes — is real, loved, and clearly monetizable, and the founder-market fit (a rocket engineer, a real-estate financier, and a finance operator who lived the customer’s pain) is better than most. The 700-customer, NPS-80 traction says the product solves a genuine problem.

But a capture layer on top of QuickBooks and Sage is a feature business, not a category. The venture case requires the 2026 agent bet to pay off: that AI can produce a construction month-end close controllers actually trust, letting Adaptive move from the invoice inbox to the system of record and expand revenue per customer several-fold. Evidence that would settle it: net revenue retention above ~120%, customers naming Adaptive (not Sage) as their book of record, and contract values climbing as modules stack. Evidence against: reviewers still routing around the accounting depth, pricing that can’t clear four figures a month, and Procore or Intuit shipping “good enough” AI coding inside software the GC already pays for. With $26.4M raised against far better-funded incumbents, Adaptive does not have unlimited runway to prove the agent story — the next raise, and its price, will tell you which business this is.

How a challenger would attack it

Attack the layer beneath it. Adaptive’s whole position rests on being the capture-and-coding layer between the invoice inbox and someone else’s system of record. A challenger doesn’t need to out-build the modules — it needs to make that layer free. Intuit or Procore can do this by bundling: ship “good enough” AI invoice coding inside software the GC already pays for, and Adaptive’s $575-$1,000/month becomes a line item to cut. A startup challenger would instead go where Adaptive is thin: payroll and union complexity (Foundation’s turf, which Adaptive “does not yet touch”) plus AP in one construction-native ledger, so the contractor gets an actual system of record instead of a sync target. The pricing confusion is a second opening — two published anchors that don’t reconcile invite a competitor with one transparent, revenue-indexed price. Third, exploit the trust gap in the agent pitch: reviewers already call the accounting depth clunky and flag vendor-credit handling, and job-costing errors are expensive. A challenger that sells “agents with a CPA-grade audit trail” — every coding decision explainable and reversible — turns Adaptive’s boldest claim into its liability with conservative contractors.

Same playbook, new buyer

The playbook — AI invoice capture coded to project, phase and cost code, then draws and WIP on top — transfers to any industry that runs job-cost accounting against progress billing. Property management and real-estate development are the nearest moves (Adaptive’s founders literally started there before pivoting away); specialty trades with union payroll, and adjacent project industries like landscaping, marine, and oil-field services all keep books in QuickBooks plus spreadsheets with no cost-code model. Geography is the other axis: the two-to-three-month payment cycle and paper-PDF workflow are worse in markets like the UK, Canada and Australia, where AIA-style US draw formats don’t apply and Adaptive’s templates don’t travel. The incumbent won’t follow fast for the same reason the wedge worked: at ~50 people and $26.4M raised against Procore, Sage and Intuit, Adaptive must concentrate everything on winning US general contractors before its runway question — “the next raise, and its price” — comes due. Fragmenting across verticals now would be strategic suicide, which is exactly what leaves the adjacent lanes open.

Sources and further reading

• TechCrunch — a16z leads $6.5M seed round for Adaptive (TechCrunch, July 2022)
• TechCrunch — Adaptive builds automation tools to speed up construction payments (TechCrunch, July 2024)
• PR Newswire — Adaptive Closes $19M Series A to Transform Construction Finance (PR Newswire, July 2024)
• Adaptive — Pricing (Adaptive, accessed July 2026)
• Adaptive — 500 Construction Companies Trust Adaptive (Adaptive, 2025)
• CPA Practice Advisor — Adaptive Unveils Project Accounting Agents (CPA Practice Advisor, June 2026)
• Capterra — Adaptive reviews, pricing and alternatives (Capterra, accessed 2026)
• Precedence Research — Construction Accounting Software Market (Precedence Research, 2025)
• Tracxn — Adaptive company profile (Tracxn, 2026)
• Emergence Capital — Adaptive portfolio page (Emergence Capital, accessed 2026)